These recommendations are based on Guidelines on the management of cellulitis, published by the Clinical Resource Efficiency Support Team [CREST, 2005]. They are pragmatic in nature, and reflect the need for thorough assessment in order to effectively manage the infection.

• A classification system has been described that grades the severity of cellulitis from class I to IV. However, only class I cellulitis should be managed in primary care; that is 'patients (who) have no signs of systemic toxicity, have no uncontrolled comorbidities, and can usually be managed with oral antimicrobials on an outpatient basis' [CREST, 2005].

These recommendations are consistent with Guidelines on the management of cellulitis, published by the Clinical Resource Efficiency Support Team on the management of cellulitis [CREST, 2005]. They are pragmatic in nature and aim to ensure that people with potentially serious cellulitis, or people who are particularly vulnerable to infection, receive the treatment they need in secondary care.

• Most people who are admitted to secondary care will require intravenous antibiotics.
• Severe localized infection, or systemic features (which may indicate the development of bacteraemia) can be life threatening (for example if they progress to necrotizing fasciitis).
• Certain groups (e.g. the very young and old, and people with comorbidities) are more vulnerable to life-threatening infection, so require a lower threshold for admission.
• People with cellulitis affecting anatomical areas where the consequence of tissue damage would be particularly critical (for example facial erysipelas) require immediate assessment.
• CKS recommends that people who do not respond to oral antibiotics, or who have frequent recurrence of cellulitis, should be referred to the appropriate department where they can:
• Be assessed further to confirm the diagnosis of cellulitis (referral to dermatology is appropriate if the diagnosis is in doubt).
• Receive intravenous or prophylactic antibiotic treatment if necessary (referral to general medicine or outpatient parenteral antimicrobial therapy service, where available, is appropriate).

These recommendations are consistent with Guidelines on the management of cellulitis, published by the Clinical Resource Efficiency Support Team (CREST) [CREST, 2005].

• There is generally a lack of evidence from placebo controlled trials to support the effectiveness of antibiotics in the treatment of cellulitis, but limited evidence from comparative trials indicates that the appropriate antibiotic will successfully treat cellulitis in about 85% of cases [Jones, 2002].
• The choice of antibiotic should be guided by known antibiotic sensitivities to the likely causative pathogens:
• Flucloxacillin is a relatively narrow-spectrum antibiotic licensed for the treatment of cellulitis [ABPI Medicines Compendium, 2008]. It demonstrates suitable pharmacokinetics, with good diffusion into skin and soft tissues [Finch et al, 2003] and, at high doses, is active against the large majority of staphylococcal and streptococcal species that cause cellulitis [CREST, 2005].
• Erythromycin is a macrolide antibiotic with a broad spectrum of activity and is suitable (and licensed) as an alternative to flucloxacillin, for example in people with an allergy to penicillin. The spectrum of activity of erythromycin includes activity against most staphylococcal species (including Staphylococcus aureus) and some Gram-negative cocci and anaerobes which may be implicated [Finch et al, 2003].
• Clarithromycin is licensed and recommended by CREST as an alternative to flucloxacillin [CREST, 2005], and has a similar spectrum of activity as erythromycin, but is reputed to cause fewer adverse effects [DTB, 1991]. The perceived superior adverse effect profile of clarithromycin compared with erythromycin is mainly theoretical, although there are some limited data from randomized controlled trials to corroborate it [Aronson, 2006].
• Co-amoxiclav is recommended by the Health Protection Agency for treatment of facial cellulitis in primary care [HPA and Association of Medical Microbiologists, 2008].
• Ciprofloxacin is recommended by CREST in addition to flucloxacillin (or a macrolide) if the cellulitis has arisen from a wound contaminated with fresh (un-chlorinated) water [CREST, 2005]. This is to cover the possibility of infection with aeromonas species [Swartz, 2004].
• Doxycycline should be considered, in addition to flucloxacillin (or a macrolide), if there has been contamination with salt water, to cover the possibility of infection with Vibrio vulnificus [Swartz, 2004].
• Clindamycin is an option recommended by the Health Protection Authority for the treatment of cellulitis in people who are allergic to penicillin [HPA and Association of Medical Microbiologists, 2008]. However, it has a worse adverse effect profile than the macrolides [BNF 55, 2008], so should be considered second-line.

These recommendations are based on Guidelines on the management of cellulitis, published by the Clinical Resource Efficiency Support Team [CREST, 2005]. In the absence of evidence from controlled trials or observational studies, self-care recommendations are based on pragmatism and clinical experience.

These are pragmatic recommendations that are consistent with Guidelines on the management of cellulitis, published by the Clinical Resource Efficiency Support Team [CREST, 2005].

• There is a lack of evidence on the optimal duration of antibiotic treatment, but clinical experience suggests that uncomplicated cases usually respond after 1–2 weeks. It is therefore prudent to prescribe a week's course of antibiotics, and follow up to ascertain whether further treatment is required. It is important to continue treatment until the condition is fully resolved.
• People should be encouraged to return if local symptoms deteriorate or systemic symptoms develop as this would put them into the referral category. However, increased erythema is sometimes reported during the first 48 hours of treatment, and is thought to be related to toxin release.